Sentinel-1 RTC Gamma0 data 
Index¶
Overview¶
This notebook demonstrates how to load and use Sentinel-1 Radiometric Terrain Corrected (RTC) Gamma0 data generated in EASI.
These analysis ready data S1 gamma-0 backscatter data are processed from Sentinel-1 GRD scenes using the SNAP-10 Toolbox with Graph Processing Tool (GPT) xml receipes. See the RTC Gamma0 product variants section for further details.
For most uses we recommend the smoothed 20 m product (sentinel1_grd_gamma0_20m).
We can process the 10 m products (sentinel1_grd_gamma0_10m, sentinel1_grd_gamma0_10m_unsmooth) and other variants on request.
Using Sentinel-1 backscatter data¶
An excellent introduction and overview to using SAR data is provided in the CEOS Laymans SAR Interpretation Guide. This guide has also been converted to a set of set of Jupyter notebooks that you can download from https://github.com/AMA-Labs/cal-notebooks/tree/main/examples/SAR.
Synthetic Aperture Radar operates in the microwave range of the electromagnetic spectrum as an active pulse sent by the satellite and scattered by features on the Earth's surface. The return signal from the surface is measured at the satellite in terms of the signal intensity, phase and polarisation compared to the signal that was sent.
The SAR instrument on the Sentinel-1 satellites operate in the C-band at approximately 5.6 cm wavelength. This means that it can "see" objects of about this size and larger, and smaller objects are relatively transparent. This makes Sentinel-1 more sensitive to tree canopies, sparse and low biomass vegetation, and surface water (wind waves or relative difference).
The SAR signal responds to the orientation and scattering from surface features of comparable size or larger than the wavelength.
- A bright backscatter value typically means the surface was orientated perpendicular to the signal incidence angle and most of the signal was reflected back to the satellite (direct backscatter)
- A dark backscatter value means most of the signal was reflected away from the satellite (forward scattering) and typically responds to a smooth surface (relative to the wavelength) such as calm water or bare soil
- Rough surfaces (relative to the wavelength) result in diffuse scattering where some of the signal is returned to the satellite.
- Complex surfaces may result in volume scattering (scattering within a tree canopy) or double-bounce scattering (perpendicular objects such as buildings or water anfd bridges)
- The relative backscatter values of co-polarisation (VV) and cross-polarisation (VH) measurements can provide information on the scattering characteristics of the surface features.
Using Sentinel-1 backscatter data requires interpretation of the data for different surface features, including as these features change spatially or in time. It may also be necessary to carefully consider the incidence angle of the SAR signal relative to the surface features using the incidence_angle band or the satellite direction metadata (descending = north to south; ascending = south to north).
Units and conversions¶
The sentinel1_grd_gamma0_* data are given in Intensity (or backscatter power) units. Intensity can be converted to decibel (dB) or amplitude, and vice-versa, with the following equations. Practical Xarray examples are given below.
Intensity to/from dB:
dB = 10 * log10(intensity) + K
intensity = 10^((dB-K)/10)
where K is a calibration factor, which for Sentinel-1 is 0 dB.
Intensity to/from Amplitude:
intensity = amplitude * amplitude
amplitude = sqrt(intensity)
Additional reference: https://forum.step.esa.int/t/what-stage-of-processing-requires-the-linear-to-from-db-command
Set up¶
Imports¶
# Common imports and settings
import os, sys, re
from pathlib import Path
from IPython.display import Markdown
import pandas as pd
pd.set_option("display.max_rows", None)
import xarray as xr
import numpy as np
# Datacube
import datacube
from datacube.utils.aws import configure_s3_access
import odc.geo.xr # https://github.com/opendatacube/odc-geo
from datacube.utils import masking # https://github.com/opendatacube/datacube-core/blob/develop/datacube/utils/masking.py
from dea_tools.plotting import display_map, rgb # https://github.com/GeoscienceAustralia/dea-notebooks/tree/develop/Tools
# Basic plots
%matplotlib inline
# import matplotlib.pyplot as plt
# plt.rcParams['figure.figsize'] = [12, 8]
# Holoviews
import hvplot.xarray
import cartopy.crs as ccrs
# EASI defaults
# These are convenience functions so that the notebooks in this repository work in all EASI deployments
# The `git.Repo()` part returns the local directory that easi-notebooks has been cloned into
# If using the `easi-tools` functions from another path, replace `repo` with your local path to `easi-notebooks` directory
try:
import git
repo = git.Repo('.', search_parent_directories=True).working_tree_dir # Path to this cloned local directory
except (ImportError, git.InvalidGitRepositoryError):
repo = Path.home() / 'easi-notebooks' # Reasonable default
if not repo.is_dir():
raise RuntimeError('To use `easi-tools` please provide the local path to `https://github.com/csiro-easi/easi-notebooks`')
if repo not in sys.path:
sys.path.append(str(repo)) # Add the local path to `easi-notebooks` to python
from easi_tools import EasiDefaults
from easi_tools import initialize_dask, xarray_object_size, heading
EASI defaults¶
These default values are configured for each EASI instance. They help us to use the same training notebooks in each EASI instance. You may find some of the functions convenient for your work or you can easily override the values in your copy of this notebook.
easi = EasiDefaults()
family = 'sentinel-1'
product = easi.product(family) # 'sentinel1_grd_gamma0_20m'
display(Markdown(f'Default {family} product for "{easi.name}": [{product}]({easi.explorer}/products/{product})'))
Successfully found configuration for deployment "asia"
Default sentinel-1 product for "asia": sentinel1_grd_gamma0_20m
Dask cluster¶
Using a local Dask cluster is a good habit to get into. It can simplify loading and processing of data in many cases, and it provides a dashboard that shows the loading/processing progress.
To learn more about Dask see the set of dask notebooks.
# Local cluster
cluster, client = initialize_dask(workers=4)
display(client)
# Or use Dask Gateway - this may take a few minutes
# cluster, client = initialize_dask(use_gateway=True, workers=4)
# display(client)
Successfully found configuration for deployment "asia"
Client
Client-3feec25c-b0f9-11ef-8346-dac6e87879d8
| Connection method: Cluster object | Cluster type: distributed.LocalCluster |
| Dashboard: https://hub.asia.easi-eo.solutions/user/pag064/proxy/8787/status |
Cluster Info
LocalCluster
87e842e7
| Dashboard: https://hub.asia.easi-eo.solutions/user/pag064/proxy/8787/status | Workers: 4 |
| Total threads: 8 | Total memory: 24.00 GiB |
| Status: running | Using processes: True |
Scheduler Info
Scheduler
Scheduler-8381929d-1cfa-46d2-b0e3-8a7776018399
| Comm: tcp://127.0.0.1:45387 | Workers: 4 |
| Dashboard: https://hub.asia.easi-eo.solutions/user/pag064/proxy/8787/status | Total threads: 8 |
| Started: Just now | Total memory: 24.00 GiB |
Workers
Worker: 0
| Comm: tcp://127.0.0.1:36925 | Total threads: 2 |
| Dashboard: https://hub.asia.easi-eo.solutions/user/pag064/proxy/35271/status | Memory: 6.00 GiB |
| Nanny: tcp://127.0.0.1:34379 | |
| Local directory: /tmp/dask-scratch-space/worker-85jovzkp | |
Worker: 1
| Comm: tcp://127.0.0.1:45123 | Total threads: 2 |
| Dashboard: https://hub.asia.easi-eo.solutions/user/pag064/proxy/37239/status | Memory: 6.00 GiB |
| Nanny: tcp://127.0.0.1:39989 | |
| Local directory: /tmp/dask-scratch-space/worker-0fdjz_6_ | |
Worker: 2
| Comm: tcp://127.0.0.1:35673 | Total threads: 2 |
| Dashboard: https://hub.asia.easi-eo.solutions/user/pag064/proxy/44647/status | Memory: 6.00 GiB |
| Nanny: tcp://127.0.0.1:34085 | |
| Local directory: /tmp/dask-scratch-space/worker-69qyhg4m | |
Worker: 3
| Comm: tcp://127.0.0.1:34201 | Total threads: 2 |
| Dashboard: https://hub.asia.easi-eo.solutions/user/pag064/proxy/37773/status | Memory: 6.00 GiB |
| Nanny: tcp://127.0.0.1:32869 | |
| Local directory: /tmp/dask-scratch-space/worker-tm3ld4rf | |
ODC database¶
Connect to the ODC database. Configure the environment and low-level tools to read from AWS buckets.
dc = datacube.Datacube()
# Access AWS "requester-pays" buckets
# This is necessary for reading data from most third-party AWS S3 buckets such as for Landsat and Sentinel-2
configure_s3_access(aws_unsigned=False, requester_pays=True, client=client);
Example query¶
Change any of the parameters in the query object below to adjust the location, time, projection, or spatial resolution of the returned datasets.
Use the Explorer interface to check the temporal and spatial coverage for each product.
# Explorer link
display(Markdown(f'See: {easi.explorer}/products/{product}'))
# EASI defaults
display(Markdown(f'#### Location: {easi.location}'))
latitude_range = easi.latitude
longitude_range = easi.longitude
time_range = easi.time
# Or set your own latitude / longitude
# Australia GWW
# latitude_range = (-33, -32.6)
# longitude_range = (120.5, 121)
# time_range = ('2020-01-01', '2020-01-31')
# Example: PNG
latitude_range = (-4.26, -3.75)
longitude_range = (144.03, 144.74)
time_range = ('2020-01-01', '2020-05-31')
# Bangladesh
# latitude_range = (21.5, 23.5)
# longitude_range = (89, 90.5)
# time_range = ('2024-05-01', '2024-06-10')
# Vietnam
# (optional target projection = epsg:32648)
# latitude_range = (9.1, 9.9)
# longitude_range = (105.6, 106.4)
# time_range = ('2024-01-01', '2024-09-10')
query = {
'product': product, # Product name
'x': longitude_range, # "x" axis bounds
'y': latitude_range, # "y" axis bounds
'time': time_range, # Any parsable date strings
}
# Convenience function to display the selected area of interest
display_map(longitude_range, latitude_range)
Location: Lake Tempe, Indonesia¶
Load data¶
# Target xarray parameters
# - Select a set of measurements to load
# - output CRS and resolution
# - Usually we group input scenes on the same day to a single time layer (groupby)
# - Select a reasonable Dask chunk size (this should be adjusted depending on the
# spatial and resolution parameters you choose
load_params = {
'group_by': 'solar_day', # Scene grouping
'dask_chunks': {'latitude':2048, 'longitude':2048}, # Dask chunks
}
# Load data
data = dc.load(**(query | load_params))
display(xarray_object_size(data))
display(data)
'Dataset size: 5.16 GB'
<xarray.Dataset> Size: 6GB
Dimensions: (time: 51, latitude: 2550, longitude: 3550)
Coordinates:
* time (time) datetime64[ns] 408B 2020-01-03T08:39:20 ... 2020-05-3...
* latitude (latitude) float64 20kB -3.75 -3.75 -3.751 ... -4.26 -4.26
* longitude (longitude) float64 28kB 144.0 144.0 144.0 ... 144.7 144.7
spatial_ref int32 4B 4326
Data variables:
vh (time, latitude, longitude) float32 2GB dask.array<chunksize=(1, 2048, 2048), meta=np.ndarray>
vv (time, latitude, longitude) float32 2GB dask.array<chunksize=(1, 2048, 2048), meta=np.ndarray>
angle (time, latitude, longitude) float32 2GB dask.array<chunksize=(1, 2048, 2048), meta=np.ndarray>
Attributes:
crs: EPSG:4326
grid_mapping: spatial_ref# When happy with the shape and size of chunks, persist() the result
data = data.persist()
Conversion and helper functions¶
# These functions use numpy, which should be satisfactory for most notebooks.
# Calculations for larger or more complex arrays may require Xarray's "ufunc" capability.
# https://docs.xarray.dev/en/stable/examples/apply_ufunc_vectorize_1d.html
#
# Apply numpy.log10 to the DataArray
# log10_data = xr.apply_ufunc(np.log10, data)
def intensity_to_db(da: 'xr.DataArray', K=0):
"""Return an array converted to dB values"""
xx = da.where(da > 0, np.nan) # Set values <= 0 to NaN
xx = 10*np.log10(xx) + K
xx.attrs.update({"units": "dB"})
return xx
def db_to_intensity(da: 'xr.DataArray', K=0):
"""Return an array converted to intensity values"""
xx = np.power(10, (da-K)/10.0)
xx.attrs.update({"units": "intensity"})
return xx
def make_image(ds: 'xarray', frame_height=300, **kwargs):
"""Return a Holoviews object that can be displayed or combined"""
spatial_dims = ds.odc.spatial_dims
defaults = dict(
cmap="Greys_r",
y = spatial_dims[0], x = spatial_dims[1],
rasterize = True,
geo = True,
frame_height = frame_height,
clabel = ds.attrs.get('units', None),
)
defaults.update(**kwargs)
return ds.hvplot.image(**defaults)
def select_valid_time_layers(ds: 'xarray', percent: float = 5):
"""Select time layers that have at least a given percentage of valid data (e.g., >=5%)
Example usage:
selected = select_valid_time_layers(ds, percent=5)
filtered == ds.sel(time=selected)
"""
spatial_dims = ds.odc.spatial_dims
return ds.count(dim=spatial_dims).values / (ds.sizes[spatial_dims[0]]*ds.sizes[spatial_dims[1]]) >= (percent/100.0)
# Examples to check that the intensity to/from dB functions work as expected
# xx = data.vv.isel(time=0,latitude=np.arange(0, 5),longitude=np.arange(0, 5))
# xx[0] = 0
# xx[1] = -0.001
# display(xx.values)
# yy = intensity_to_db(xx)
# display(yy.values)
# zz = db_to_intensity(yy)
# display(zz.values)
# Optional time layer filter
selected = select_valid_time_layers(data.vv, 10)
data = data.sel(time=selected).persist()
/env/lib/python3.12/site-packages/rasterio/warp.py:387: NotGeoreferencedWarning: Dataset has no geotransform, gcps, or rpcs. The identity matrix will be returned. dest = _reproject( /env/lib/python3.12/site-packages/rasterio/warp.py:387: NotGeoreferencedWarning: Dataset has no geotransform, gcps, or rpcs. The identity matrix will be returned. dest = _reproject( /env/lib/python3.12/site-packages/rasterio/warp.py:387: NotGeoreferencedWarning: Dataset has no geotransform, gcps, or rpcs. The identity matrix will be returned. dest = _reproject( /env/lib/python3.12/site-packages/rasterio/warp.py:387: NotGeoreferencedWarning: Dataset has no geotransform, gcps, or rpcs. The identity matrix will be returned. dest = _reproject(
# Add db values to the dataset
data['vh_db'] = intensity_to_db(data.vh).persist()
data['vv_db'] = intensity_to_db(data.vv).persist()
Plot the data¶
Note the different data ranges for plotting (
clim) betweenvv,vh, intensity and dB.
- Stronger co-polarisation (VV) indicates direct backscatter while stronger cross-polarisation (VH) may indicate a complex surface or volume scattering.
- Intensity data are linear-scaled so can tend to disciminate across a range of backscatter returns.
- Decibel data are log-scaled so can tend to discriminate high and low backscatter returns.
# A single time layer for VV and VH, with linked axes. Adjust browser window width if required.
idx = 0
time_label = data.time.dt.strftime("%Y-%m-%d %H:%M:%S").values[idx]
vvplot = make_image(data.vv.isel(time=idx), clim=(0, 0.5), title=f'VV ({time_label})')
vhplot = make_image(data.vh.isel(time=idx), clim=(0, 0.1), title=f'VH ({time_label})')
angle = make_image(data.angle.isel(time=idx), title=f'Angle ({time_label})')
vvplot + vhplot + angle
# Make a dB plot. Adjust browser window width if required.
vvplot = make_image(data.vv_db.isel(time=idx), clim=(-30, -3), title=f'VV ({time_label})')
vhplot = make_image(data.vh_db.isel(time=idx), clim=(-30, -1), title=f'VH ({time_label})')
vvplot + vhplot + angle
# Subplots for each time layer for VV, with linked axes
make_image(data.vh_db, clim=(-30, -3), robust=True).layout().cols(4)